I would support the hell out of variable weather, but the technical difficulties associated with implementing it is incredible.
First off, a new map would have to be made for the weather-varied map. For example, we'll take Skidrow. Let's add rain to Skidrow.
We first begin by adding water-type textures around the map. Puddles, changing dirt to mud, grass to wet grass, etc. This requires tedious time and effort, as well as a seperate map and script associated with it (maps cannot share scripts, they're all uniquely named: instead of skidrow.map it would be skidrow_wet.map for example). You've now doubled the space for one map that the game has to hold. This slows down installation speeds and restricts space for other things.
In addition, you have to write the code for rain. This is easy enough as long as you have to do it, bear in mind, you have to create another script for this as well, so you're doubling the space that would have to be used to store this.
All in all, this is just economically unfriendly and unnecessary. It's a good idea in theory, a massive headache in implementation and it will never happen to call of duty unless the engine is modified to allow for it - but why?
And no, you can't use the same map because it would be almost impossible to specify all the changed textures to script_brushmodels and delete/add them via scripts - okay, well, it's possible, but very tedious, very unnecessary, and most likely, very prone to failure and glitches.
A day/night cycle would be easier, but just as bad (you'd have to duplicate the map and then change its skybox/worldspawn (sunlight) settings). But this suffers the same problem. In the end, it's not going to happen.
If you actually read this, congrats. tl;dr not economical/no easy solution/space issues
